Brain-wide mapping of fMRI network dynamics in the mouse brain

Intrinsic brain activity has been widely characterized using the blood-oxygen-level-dependent(BOLD) functional Magnetic Resonance Imaging(fMRI)at rest. There is increasing interest in finding reproducible and robust signatures of large-scale brain synchronization, and pinpointing their neurophysiological substrates and inherent alteration in disease.In this respect, the implementation of dynamic fMRI mapping in laboratory animalsrepresents a major advance, offeringthe opportunity to unravel the elusive drivers of this phenomenon via the use of cell-type specific manipulations that are off limits in humans.Multiple investigations have shown that spontaneous brain activity is non-stationary andinvolves reconfigurationsinto multiple dynamicstates.This research describesa series of studies aimedto map spontaneous fMRI (rsfMRI) network dynamics in the resting mouse brain with voxel resolution. Starting from a proof-of-concept demonstration that canonical resting state fMRI correlations are reliably described by brief instances of regional peak fMRI activity, wedevised a novel frame-wise clustering strategy that allowed us to map recurrent fMRI networks states dynamicsin the mouse brain. We showthat brain-wide patterns of fMRI co-activation can be reliably mapped at the group and subject level, defining a restricted set of recurring brain states characterized by rich network structure. Of particular interest was the observation of opposite co-activation of the mouse default mode network (DMN) and Latero-cortical networks(LCN), two systems that have been proposed to parallel analogous systems of the human brain.Importantly, we also document that these functional states are characterized by contrasting patterns of spontaneous fMRI activity,and exhibit coupled oscillatory dynamics embedded in a common temporal reference marked by infra-slow global fMRI signal oscillations. We next applied this novel framework to a genetic modelof autismand show that aberrant patterns of fMRI connectivity in a genetic model of autism reflect the engagement non-canonical brain states, characterized by altered regional topography and oscillatory dynamics. We finally show that pharmacological stimulation of the cholinergic systems results in reduced large-scale brain synchronization, a finding associated with anew set of oscillating statesin which the involvement of basal forebrain areas is pre-dominant. Collectively,our result demonstrate the possibility of mapping spatio-temporal dynamics of spontaneous brain activity in the living mouse brain with voxel resolution. Our approach reveals a new set of fundamental principles guiding the spatiotemporal organization of resting state fMRI activity, and its disruption in brain disorders.